On meson spectral functions at high temperature and nonzero momentum

PROCEEDINGS

OF SCIENCE

On meson spectral functions at high temperature
and nonzero momentum

Gert Aarts*f Simon Hands 5

University of Wales Swansea

E-mail: |g . aart s@ swan .ac.uk s.hands@swan.ac.uk

Seyong Kim

Sejong University, Seoul
E-mail:

skim@ sejong. ac . kr

Jose M. Martinez Resco

Brandon University
E-mail: jnart inez resco jtjjbrandonu . ca

In the high-temperature phase of QCD meson spectral functions at nonzero momentum are ex-
pected to have a nontrivial and interesting structure. In order to provide a reference point for lattice
studies employing e.g. the Maximal Entropy Method, we discuss several characteristics of meson
spectral functions in the infinite-temperature limit. We report on ongoing work in quenched QCD
with staggered fermions.

XXIIIrd International Symposium on Lattice Field Theory

25-30 July 2005

Trinity College, Dublin, Ireland

* Speaker.

*PPARC Advanced Fellow
§ PPARC Senior Research Fellow

© Copyright owned by the author(s) under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike Licence.

http: //pos . sissa . it /

On meson spectral functions at high temperature and nonzero momentum

Gert Aarts

1. Spectral functions

The recent experimental progress in the recreation of the quark gluon plasma in relativistic
heavy ion collisions (see e.g. [|l|]) has been an important stimulus for lattice studies of spectral
functions in the high-temperature phase of QCD. Topics discussed so far (ordered according to
increasing difficulty) include the persistent presence of charmonium bound states above the decon-
finement transition ^ ^J, the thermal dilepton rate [j|], and transport coefficients, such as the
electrical conductivity [^] and the shear viscosity In spectral function calculations soft energies
CO < T are of particular interest, since this is where one expects e.g. nonperturbative medium ef-
fects and collective excitations, predicted by perturbative studies of hot QCD. Moreover, transport
coefficients follow from current-current spectral functions in the limit of vanishing energy (Kubo
relations). So far numerical results for the soft energy region in the high-temperature phase are
rather scarce. For the dilepton rate, it was found (surprisingly) that the spectral function in
the vector channel p v (ft),0) is consistent with zero when CO < 2T. For the electrical conductivity,
p"(<o,0) was reconstructed at low energies ^ and a structure resembling analytical expectations
[^] was obtained.

However, numerically determined euclidean correlators G//(T,p) do not easily provide knowl-
edge of spectral functions at soft energies. In fact, they are largely insensitive to details of p#(ft),p)
in this region because the kernel K, which relates G# and pn via

„ / \ f°° d(0 . . . . cosh[o(T-l/2r)]
G " (T ' P)= /o 2^ K ^°»P«^ K ^)= smh{co/2T) » (I'D

becomes independent of x for smaller energies (K(r,co) ~ 2T/co when CO <^ T). It follows in
particular that lattice conelators are remarkably insensitive to transport coefficients [||] (see ref.
[^] for a recent analysis reaching the same conclusion). In view of the experimental results on the
strongly interacting QGP, suggesting e.g. a surprisingly small shear viscosity, it is important to gain
more experience in the reconstruction of spectral functions at energies CO <T from lattice QCD.

Good candidates for such studies are meson spectral functions at nonzero momentum p JTo|] -
Because of the presence of the lightcone at CO = p and the nontrivial spectral weight below the
lightcone, commonly referred to as Landau damping and due to the scattering of quarks with off-
shell gauge bosons, they are expected to have an interesting structure at soft energies. By choosing
the nonzero momentum p of the order of T or larger, the region below the lightcone can be made
sufficiently large and (part of) the difficulties present for transport coefficients, inherent in the strict
limit CO — > 0, can be circumvented. In order to provide a reference point for such calculations, we
studied meson spectral functions at nonzero momentum in the infinite temperature limit, both in

the continuum and on the lattice for Wilson and staggered (naive) fermions JlOfl- A similar study
at zero momentum can be found in ref. [pi]]. Below we summarize some of our findings. In the
outlook we report on ongoing work in quenched QCD with staggered fermions.

2. Infinite temperature

We consider meson spectral functions p#(/,x) = x), 7^(0,0]) at leading order in the

loop expansion. The currents in the various channels are given by Jjj(t,x) = q(t,x)rjiq{t,x) with

182/2

On meson spectral functions at high temperature and nonzero momentum

Gert Aarts

r# = { 1 , 75 , y' 1 , 7^ 75 }■ In the continuum the one-loop integral

-— T tr5(jffl/ J k)rH5(^ + jftj/,p + k)7°47 , (2.1)

where 5(j£0/,k) is the fermion propagator, can be done analytically, for arbitrary external momen-

tum p and fermion mass m. Since the outcome is rather lengthy, we refer to ref. Q10| ] for detailed
expressions. On the lattice completely analytical results are not available, but simple expressions
which are easily evaluated numerically can be given. In order to do so, we use a 'mixed' represen-
tation Jl2[ ] and write the fermion propagator as 5(r,k) = 7454(1, k) +£|=i 7''S,-(T,k) + 15 M (T,k).
For Wilson fermions, the 5(t,k) functions depend on x via sinh(f£'| 4 ) and cosh(f£k), where
X = X — 1/2T, which makes it straightforward to arrive at the relationship ( |1.1| ), also on a finite
lattice. The resulting spectral functions are then given by a sum over the spatial lattice momenta,
which can be done numerically.

For naive fermions there are a few changes, due to the time doublers. The hyperbolic functions
are now multiplied with the staggering factors 1 ± (— 1) T /" T , such that relation ( |1.1[ ) is modified to

G H (x,p) = 2 —K(x,(o) PH ((0,p)-(-iy^p H (o),p), (2.2)

JO 27T L J

where K is again the same kernel as in the continuum. The staggered partner pn is related to p#
via the replacement T# — > f# = T^sT//.

We now discuss several characteristics. It is expected from naive dimensional arguments that
meson spectral functions increase with ft) 2 for large ft). This is demonstrated in fig. [I] (top), where
we show the (pseudo)scalar spectral functions pps,s(^,p), normalized with ft) 2 , in the continuum
and on the lattice for finite Af T = 16. In order to approximate the thermodynamic limit, N a is
taken very large, typically ~ 2000 Jl0|]. We only show results for isotropic lattices here, results
for anisotropic lattices can be found in ref. JTOj] . The cusps at larger energies are due to the finite
Brillouin zone and are therefore lattice artefacts. They are located at ft) = £"k-p/2 + ^k+p/2 ~ 2£"k,

Wilson staggered

20 40 60 20 40 60

OJ/T

Figure 1: Pseudoscalar and scalar spectral functions pps,s(ft>,p)/a) 2 (top) and charge density spectral func-
tions p 00 (co, p)/r 2 (bottom) as a function of co/T.

182/3

On meson spectral functions at high temperature and nonzero momentum

Gert Aarts

with k = (n/a,0,0),(n/a,7t/a,0) + permutations for Wilson fermions. The maximal energy is
determined by k = (n/a,n/a,n/a) (k= (n /2a, K /2a, 71 /2a) for staggered fermions). For Wilson
fermions pps and ps differ, due to the presence of the Wilson mass term, while for staggered and
continuum fermions they are degenerate for larger energies.

Charge density spectral functions do not increase with ft) 2 , but instead reach a constant value:
p 0() (ft),p) — > N c p 2 /6% (vector charge density) and pg°(ft),p) — > N c (p 2 + 6m 2 )/6% (axial charge
density). Note that this is relevant for the choice of default model in the Maximal Entropy Method
Jl3|]. For the vector charge density, this large energy behaviour follows from current conserva-
tion. In fact, at zero momentum current conservation completely fixes the charge density spectral
function to be p 00 (ft),0) = 2n%(o8{(o), where % is the charge susceptibility. The corresponding
euclidean correlator is then constant, G 00 (t,0) = T%. On the lattice these expressions only hold
when the conserved current is used, and not the local one. The continuum and lattice spectral func-
tions p 00 are compared in fig. [j] (bottom) for nonzero p = AT. Note that the local definition is used
here. The staggered result appears to track the continuum one up to larger energies than the one
obtained with Wilson fermions.

Clearly visible in fig. [j] is the lightcone is at ft) = p = AT and the spectral weight below the
lightcone. In the (pseudo)scalar channel, a better way to view this physically interesting region is
by normalizing pps,s(^,p) with T 2 instead of with ft) 2 . The result is shown in fig. |2] (top). The
spectral functions increase linearly with ft) for small ft) and vanish at the lightcone. There is a gap
when p < ft) < y p 2 + 4m 2 (higher loop corrections will fill this gap). The scalar and pseudoscalar
spectral functions differ due to the nonzero quark mass. Again the staggered result compares better
with the continuum one. The main lattice artefact in this region is the mismatch between the
continuum and the lattice lightcone, which can be quite substantial.

A very convenient way to present both the low and the high energy behaviour of spectral
functions in one figure is to show p#(ft),p)/sinh(ft)/2r), i.e. the integrand in eq. ( |Q| ) at the

182/4

On meson spectral functions at high temperature and nonzero momentum

Gert Aarts

•--•71 m/T =3

a— a a m/T =3
*-»p m/T =3
• -•71 m/T=1.2
[>-□ o m/T =1.2
p m/T=1.2
»--»nm/T=0.24
°--vom/T=0.24
±-Apm/T=0.24

48 x24
(3=6.5

C3

0.1

0.01

0.001

0.0001

le-05

7C m/T =3
a--uG m/T =3
i-ip m/T=3
• • 71 m/T =7.2

o m/T =7.2
«--»p m/T=1.2
•--•nm/T=0.24
°--aam/T=0.24
^-±pm/T=0.24

64 x24
(3=7.192

«.--.

24

Figure 3: Euclidean correlators in the pseudoscalar (tt), scalar (a) and local vector (p, averaged over
/ = 1,2,3) channel for three values of the staggered fermion mass m (in units of the temperature) in quenched
QCD on a lattice with T w 160 MeV (top) and T w 420 MeV (bottom).

midpoint % = 1 /27\ This combination takes a finite value when CO — ► and vanishes exponentially
for large O). We show pv = p" — p 00 and Pa = p l 5 ' — Pj in fig. || (bottom). The lattice artefacts
at large CO are now exponentially suppressed, which implies that the euclidean correlator around
T = 1/2T is not very sensitive to those. Finally, the area under the two curves is identical, which
follows from a surprising relation ("sum rule") at this order in the loop expansion [|10|].

3. Outlook

In Fig. |3| we show preliminary results from a quenched simulation with staggered lattice

182/5

On meson spectral functions at high temperature and nonzero momentum

Gert Aarts

fermions. To facilitate comparison of hadron correlators at differing temperatures, the lattice pa-
rameters have been chosen to reproduce temperatures both below (j3 = 6.5, 48 3 x 24) and above
(jS = 7.192, 64 3 x 24) the deconfining temperature with the same number of temporal spacings
N r = 24. The data shown are for zero spatial momentum; comparison of the % and a-propagators
between the two sets, particularly for the lightest mass m/T = 0.24 (shown in red), show clear
evidence for chiral symmetry restoration in the deconfined phase. A spectral function analysis of
these data using the Maximal Entropy Method is in progress.

Acknowledgements

S.K. thanks PPARC for support during his visit to Swansea in 2004/05.

References

[1] K. Adcox et al. [PHENIX Collaboration], Formation of dense partonic matter in relativistic nucleus
nucleus collisions at RHIC: Experimental evaluation by the PHENIX collaboration, Nucl. Phys. A
757 (2005) 184 [ nucl-ex/ 04 10003]; J. Adams et al. [STAR Collaboration], Experimental and
theoretical challenges in the search for the quark gluon plasma: The STAR collaboration's critical
assessment of the evidence from RHIC collisions, ibid 757 (2005) 102 [nucl -ex/050100 9].

[2] M. Asakawa and T. Hatsuda, J /y and r\ c in the deconfined plasma from lattice QCD, Phys. Rev. Lett.
92 (2004)012001 [hep-lat/0308034].

[3] S. Datta, F. Karsch, P. Petreczky and I. Wetzorke, Behavior of charmonium systems after
deconfinement, Phys. Rev. D 69 (2004) 094507 [hep-lat/0312037].

[4] T. Umeda, K. Nomura and H. Matsufuru, Charmonium at finite temperature in quenched lattice QCD,
Eur. Phys. J. C 39S1 (2005) 9 [hep-lat / 02 1 1 3].

[5] F. Karsch, E. Laermann, P. Petreczky, S. Stickan and I. Wetzorke, A lattice calculation of thermal
dilepton rates, Phys. Lett. B 530 (2002) 147 [hep-lat/0110208].

[6] S. Gupta, The electrical conductivity and soft photon emissivity of the QCD plasma, Phys. Lett. B 597
(2004) 57 [hep-lat/030100 6].

[7] A. Nakamura and S. Sakai, Transport coefficients of gluon plasma, Phys. Rev. Lett. 94 (2005) 072305

[hep-lat/0406009].

[8] G. Aarts and J. M. Martinez Resco, Transport coefficients, spectral functions and the lattice, JHEP
0204 (2002) 053 [hep-ph/ 02 03177]; Transport coefficients from the lattice?, Nucl. Phys. Proc.
Suppl. 119 (2003) 505 [hep-lat/0209033].

[9] P. Petreczky and D. Teaney, Heavy quark diffusion from the lattice, hep-ph/ 5 07 318.

[10] G. Aarts and J. M. Martinez Resco, Continuum and lattice meson spectral functions at nonzero
momentum and high temperature, Nucl. Phys. B 726 (2005) 93 [hep-lat / 0507004]

[11] F. Karsch, E. Laermann, P. Petreczky and S. Stickan, Infinite temperature limit of meson spectral
functions calculated on the lattice, Phys. Rev. D 68 (2003) 014504 [hep-lat/0303017].

[12] D. B. Carpenter and C. F. Baillie, Free Fermion Propagators And Lattice Finite Size Effects, Nucl.
Phys. 5 260(1985) 103.

[13] M. Asakawa, T. Hatsuda and Y. Nakahara, Maximum entropy analysis of the spectral functions in
lattice QCD, Prog. Part. Nucl. Phys. 46 (2001) 459 [hep-lat / 1 1 4 0].

182/6